Open Letter January 2016

Dear Robotics Community,

The best wishes for a happy and successful new year from the RSS foundation! 

Please find below some information about the evolution of RSS and the activities of the foundation.

Best regards

The RSS Foundation Board

 

RSS Attendance Over the Years

RSS has established itself as a well-attended single-track conference in robotics. RSS 2014 in Berkeley was attended by an incredible 877 people.  This record attendance certainly was an anomaly, probably due to the large number of robotics labs and robotics conference home to the Bay Area.  At RSS 2015 in Rome, the attendance was back to 351, roughly at the levels attained for previous RSS.  One of the strong features of RSS is the ratio of attendees to papers; it is about five to seven attendees per paper.

Attendance since 2005

Foundation Financial Report & Registration Fees

The RSS foundation is a non-profit organization.  The financial policy of the foundation is to accumulate a surplus roughly corresponding to the total budget of one RSS conference.  This goal is to have sufficient funds to relieve local arrangements chairs of the financial risk they take when hosting a conference.  To organize RSS, it might be necessary to make financial commitment. Should RSS for some reason not take place (remember SARS?), the foundation will be able to over the financial loss.

The budget of an RSS conference currently amounts to approximately 200K US$.  Over the past years, the foundation has accumulated a surplus of about 170K US$.  It is possible that the conference budgets increase in future years.  Also, the profitability of a particular conference depends to a  significant degree on the level of company sponsorship raised by the organizing team.  Accounting for these factors, once the accumulated surplus has reached the level of a conference budget at the time, the RSS foundation will begin to lower the registrations fees until the conference budget breaks even.

This has always been the plan of the RSS foundation.  At the board meeting in Rome, the board voted to implement it.

Analysis of RSS Performance

In 2014, the board asked Gaurav Sukhatme to begin a longitudinal analysis of RSS performance, and in particular to determine appropriate metrics for the conference and identify possible goals. In 2015, Gaurav provided some preliminary analysis, the bulk of which is available here.

While it is too early to draw any conclusions about the performance of RSS, it is reassuring that some simple analysis indicates that RSS is having impact in terms of citations and h-index. Nevertheless, because of its size and longevity, ICRA continues to be the leader in these measures.

The board thanks Gaurav for launching this work and continuing it into the future.

RSS Chair and Site Selection

Over the last two years, the RSS foundation has worked to create a more transparent set of processes for identifying the next program and local arrangements chairs and location of the conference. The highlights include:

  • We solicit nominations for program chair and proposals for hosting RSS two years ahead of time, in advance of the board meeting at each RSS.
  • The board votes on the nominations at the board meeting, or may choose to solicit additional nominations or proposals.
  • The RSS board values diversity of various kinds, including geographic diversity, and also pays particular attention to proposals that minimise attendance costs, e.g., by providing low-cost housing solutions for students.

As of 2015, the program chair and locations for future RSS conferences are:

  • 2016: Nancy Amato program chair, Ryan Eustice and Edwin Olson local arrangements chairs, University of Michigan
  • 2017: Siddhartha Srinivasa program chair, Sanem Sarıel, Gökhan İnce and Sertaç  Karaman local arrangements chairs, Istanbul Technical University, Turkey

Proposals to host future conferences are welcome from any location and at any time preferably two years prior to the RSS conference. Guidance for how to prepare and submit nominations and proposals can be found at the foundation board website.

The RSS Dual Submission Policy and Arxiv

The nature of scientific publication has evolved over time, and the RSS foundation has recognized the need to evolve its own standards too. The desire to make technical results widely accessible in a timely manner has historically always been at odds with the need to ensure that results are rigorously reviewed before being accepted into our common body of knowledge. The popularity of new electronic publications such as arXiv raises questions of what constitutes the definitive publication of a result, and what papers can RSS accept as original work. Should RSS review submissions that have been uploaded to arXiv and already discussed  on Slashdot, Reddit and the BBC?

It is easy to assume that this is a new question for the scientific community, but the problem with “peer review by press release” predates arXiv by decades. In 1969, the New England Journal of Medicine instituted the “Ingelfinger rule”, which stated that only papers would be accepted for publication that “have been neither published nor submitted elsewhere (including news media and controlled-circulation publications)” (Ingelfinger, 1969). The author of this rule, Franz Ingelfinger, freely admitted that the rule was instituted to protect the newsworthiness of the venue, but had the secondary property that it ensured that scientific results underwent peer review. The rule is frequently criticised and has been periodically re-examined by the NEJM itself and other publications about every 10 years (Relman, 1981; Angell and Kassirer, 1991, Toy 2002), but the rule does put into stark relief the tension between getting our work out into the world and holding ourselves up to rigorous evaluation. RSS does not have anywhere near as stringent a rule as the Ingelfinger rule, but we do want a publication policy that  emphasizes innovation and impact, and we want to avoid accepting papers that are a rehash of already-known results.

Further complicating the matter for pre-tenure academics is how our publications are viewed at promotion time. Many engineering fields look dubiously on the practice in computer science of using conferences as our primary publication venue. Publications at conferences that also allow submission to other archival venues are generally ignored by most academic promotion  committees, and sometimes conferences as a whole are ignored, hence the recent RAS Letters journal.

At board meetings in 2014 and 2015, the RSS foundation discussed whether RSS should change its publication model, including rolling submissions and acceptances and how to ensure RSS papers are recognized as definitive publications. We do not plan any major changes in the near term and we will continue to have a submission policy that disallows dual submissions to other archival venues. However, we recognize the role that arXiv specifically is playing in computer science, and we will officially accept papers that also appear in arXiv.

The RSS foundation will continue to regularly evaluate our submission policy; comments and suggestions are very welcome at any time.

  1. Ingelfinger, “Definition of Sole Contribution”, N Engl J Med; 281:676-677, September 18, 1969.
    Arnold S. Relman, “The Ingelfinger Rule”, N Engl J Med; 305:824-826, October 1,1981.
    Marcia Angell and Jerome P. Kassirer. “The Ingelfinger Rule Revisited”, N Engl J Med; 325:1371-1373 November 7, 1991.
    Jennifer Toy, “The Ingelfinger rule: Franz Ingelfinger at the New England Journal of Medicine 1967-77.” Sci Ed 25(6):195-198, 2002.